Computational models of image representation in visual cortex
Each waking moment, our brain is bombarded by sensory information, estimated to be in the range of hundreds of megabits/sec. Somehow, we make sense of this
data stream by extracting the forms of spatiotemporal structure embedded in it, and from this we build meaningful representations of objects, sounds, surface
textures and so forth in the environment. The overarching goal of research in my laboratory is to understand how this process occurs in the brain, focusing
especially on the thalamo-cortical system.
Our work is based upon the hypothesis that the cortex essentially contains a probabilistic, causal model of the environment, and that sensory information is
interpreted and represented in terms of this model. Thus, one major line of work is to develop probabalistic models of natural images, and to construct neural
circuits capable of representing images in terms of these models. For example, we have developed a model of natural images based on the principle of sparse coding —
in which the retinal image is explained in terms of a small number of events at any given point in time — and we have shown that the receptive field properties
that emerge in such a system match those found in the primary visual cortex (V1) of mammals. The suggestion then is that V1 may be operating, at least in part, according
to a similar principle. We are currently working on extending this model to learn invariances from natural image sequences, in addition to building models composed of
multiple layers to capture the hierarchical structure of visual cortex.
Another line of work in our lab is to test the predictions of these models in psychophysical and neurophysiological experiments, oftentimes in collaboration with other
labs. Together with Dr. Charles Gray at Montana State University, Bozeman, we are investigating the joint activity of V1 neurons in response to natural movies in an
attempt to test certain aspects of the sparse coding model. We have also been using methods of EEG and fMRI, in addition to behavioral measures, to investigate both the
time-course and locus of object recognition and scene analysis processes occuring in human cortex. The results of these studies provide important constraints for building
computational models.
Selected Publications
Olshausen BA, Field DJ (2005) How close are we to understanding V1? Neural Computation, 17, 1665-1699.
Olshausen BA, Field DJ (2004) Sparse coding of sensory inputs. Current Opinion in Neurobiology, 14, 481-487.
Johnson JS, Olshausen BA (2003) Time-course of neural signatures of object recognition. Journal of Vision, 3, 499-512.
Olshausen BA (2003) Principles of image representation in visual cortex. In: The Visual Neurosciences, L.M. Chalupa, J.S. Werner, Eds. MIT Press. pp. 1603-15.
Olshausen BA (2002). Sparse codes and spikes. In: R.P.N. Rao, B.A. Olshausen, M.S. Lewicki (Eds.), Probabilistic Models of Perception and Brain Function, pp. 257-272. MIT Press.
Murray SO, Kersten D, Olshausen BA, Schrater P, Woods DL (2002) Shape perception reduces activity in human primary visual cortex. Proceedings of the National Academy of Sciences, USA, 99(23), 15164-15169.
Simoncelli EP, Olshausen BA (2001). Natural image statistics and neural representation. Annual Reviews of Neuroscience, 24, 1193-1215.
Links
Optometry Web Page
Helen Wills Neuroscience Institute Web Page
 Back to Faculty Directory page
|
